How to Install Tensorflow With Cuda in GPU Windows

How to Install Tensorflow With Cuda in GPU Windows with Git Bash

- install Anaconda and Git Bash

- create environment with python 3.10.13

- Setup Conda in Git Bash

dell@AHMADROIHAN-DELL MINGW64 /e/Program-Files/anaconda3/etc/profile.d
$ echo ". ${PWD}/conda.sh" >> ~/.bashrc
dell@AHMADROIHAN-DELL MINGW64 /e/Program-Files/anaconda3/etc/profile.d
$ echo ". '${PWD}'/conda.sh" >> ~/.bashrc

- Set up Git Bash to always run as Administrator

- Click the Windows Start button and type bash.

- Right-click the found link and select Open file location.

- Right-click the menu shortcut and select Properties.

- On the Compatibility tab select Run this program as administrator.

- Reopen Git Bash in other project directory and activate environment 

dell@AHMADROIHAN-DELL MINGW64 /f/PYTHON/gpu
$ conda activate p31013
(p31013)
dell@AHMADROIHAN-DELL MINGW64 /f/PYTHON/gpu
$ conda install -c conda-forge cudatoolkit cudnn

.................

dell@AHMADROIHAN-DELL MINGW64 /f/PYTHON/gpu

$ pip install "tensorflow<2.11"

.................

dell@AHMADROIHAN-DELL MINGW64 /f/PYTHON/gpu
$ python -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"
2023-12-28 06:46:05.405858: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX AVX2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2023-12-28 06:46:07.065071: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1616] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 1654 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 3050 Laptop GPU, pci bus id: 0000:01:00.0, compute capability: 8.6
tf.Tensor(1.7581177, shape=(), dtype=float32)
(p31013)
dell@AHMADROIHAN-DELL MINGW64 /f/PYTHON/gpu
$ python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"
[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]